Matplotlib drawing interface settings

Coordinate limits and titles

Code	meaning	Parameters
plt.xlim()	Definition x Axis coordinate limit	Leftmost value , The rightmost value
plt.ylim()	Definition y Axis coordinate limit	The lowest value , The top value
plt.axis()	Set coordinate limits	[xmin,xmax,ymin,ymax]
plt.axis('tight')	Tighten the axis , Leave no blank
plt.axis('equal')	Set the resolution of the graphics displayed on the screen （ Ratio of unit length of two axes ）
plt.title()	Graphic title
plt.xlabel()	x Axis title
plt.ylabel()	y Axis title
plt.style.use()	Table style
ax.spines['top'].set_color('none')	Hide the axis top boundary
ax.spines['right'].set_color('none')	Hide the right boundary of the coordinate axis
ax.xaxis.set_major_locator(MultipleLocator(2))	Definition x The scale unit of the coordinate axis	Need to be from matplotlib.pyplot import MultipleLocator
ax.yaxis.set_major_locator(MultipleLocator(0.1))	Definition y The scale unit of the coordinate axis	Need to be from matplotlib.pyplot import MultipleLocator
plt.xticks()	Change the scale	coordinates , Substitute data （[0.2, 0.4, 0.6], ['A', 'B', 'C']）

Table style

Use plt.style.available You can see all the styles

Solarize_Light2	_classic_test_patch	bmh	classic	dark_background
fast	fivethirtyeight	ggplot	grayscale	seaborn
seaborn-bright	seaborn-colorblind	seaborn-dark	seaborn-dark-palette	seaborn-darkgrid
seaborn-deep	seaborn-muted	seaborn-notebook	seaborn-paper	seaborn-pastel
seaborn-poster	seaborn-talk	seaborn-ticks	seaborn-white	seaborn-whitegrid
tableau-colorblind10

How to use style sheets

plt.style.use('fivethirtyeight')

This will change the style of all tables , if necessary , You can use the style context manager to temporarily change to another style ：

with plt.style.context('fivethirtyeight'):
    plt.plot([1,2,3], [3,1,2])

dark_background style

plt.style.use('dark_background')

x = np.linspace(0,10,1000)
fig, ax = plt.subplots()
ax.plot(x, np.sin(x),'-b')
ax.plot(x, np.cos(x), '--r')

Words and notes

plt.text()： Add notes ( be equal to ax.text())
This method requires x Axis position 、y Axis position 、 character string 、 And some optional parameters , Such as the color of the text 、 Font size 、 style 、 Alignment, etc

plt.plot([1,2,3,4,5,6], [2,5,6,1,3,4])

#  Add text to the diagram 
ax.text(1,2,(1,2), ha='center')
ax.text(3,6,'(3, 6)', ha='right')
ax.text(4,1,str((4,1)))

transform Parameters ： Coordinate transformation and text position

ax.transData： Coordinate transformation based on data （ Axis ）
ax.transAxes： Coordinate transformation based on coordinate axis （ In axis dimensions ）（ Coordinate system scale ）
fig.transFigure： Coordinate transformation based on graphics （ In drawing dimensions ）（ Figure scale ）

ax.set_xlim(0,10)
ax.set_ylim(0,10)

#  Add text to the diagram 
ax.text(1, 5, ". Data: (1, 5)", transform=ax.transData)
ax.text(0.5, 0.1, ". Axes: (0.5, 0.1)", transform=ax.transAxes)
ax.text(0.4, 0.4, ". Figure: (0.4, 0.4)", transform=fig.transFigure)

When changing the coordinate axis , Only ax.transData Your point will change

ax.set_xlim(-2,2)
ax.set_ylim(-6,6)

Arrows and notes

plt.annotate()： Draw arrows and notes

ax.set_xlim(-2,5)
ax.set_ylim(-6,6)

ax.annotate('A', xy=(3, 1), xytext=(4, 4), arrowprops=dict(facecolor='black', shrink=0.05))
ax.annotate('B', xy=(1, 1), xytext=(4, 3), arrowprops=dict(arrowstyle="->", connectionstyle="angle3,angleA=0,angleB=-90"))

ax.annotate('B', xy=(1, 1), bbox=dict(boxstyle="round",fc="none",ec="gray"), xytext=(4, 3),
            ha='center',arrowprops=dict(arrowstyle="->", connectionstyle="angle3,angleA=0,angleB=-90"))

Custom coordinate scale

Define the scale unit of the coordinate axis

ax.xaxis.set_major_locator(MultipleLocator(0.2))
ax.yaxis.set_major_locator(MultipleLocator(0.3))

Hide the upper boundary right boundary

ax.spines['top'].set_color('none')
ax.spines['right'].set_color('none')

Change the scale

plt.xticks([0.2, 0.4, 0.6, 0.8], ['A', 'B', 'C', 'D'])

Major and minor scales

The main scale tends to be larger , Secondary scales tend to be smaller , For example, logarithmic graph

#  Create graphics 
fig = plt.figure()
#  Axis 
ax = plt.axes(xscale='log', yscale='log')
ax.set_xlim(10**0,10**5)
ax.set_ylim(10**0,10**5)

Set the formatter and locator Custom scale properties

Hide scales and labels

Hidden scales and labels are usually used plt.NullLocator() And plt.NullFormatter() Realization
Below we remove X Axis labels （ But the tick marks are preserved / Gridlines ）,Y Axis scale （ The label is also removed ）

#  Create graphics 
fig = plt.figure()
#  Axis 
ax = plt.axes()
ax.set_xlim(0,5)
ax.set_ylim(0, 5)

ax.yaxis.set_major_locator(plt.NullLocator())
ax.xaxis.set_major_formatter(plt.NullFormatter())

Increase or decrease the number of scales

adopt plt.MaxNLocator() Set the maximum number of scales displayed

fig, ax = plt.subplots(4, 4, sharex=True, sharey=True)
for axi in ax.flat:
    axi.xaxis.set_major_locator(plt.MaxNLocator(5))
    axi.yaxis.set_major_locator(plt.MaxNLocator(5))

Summary of format generator and locator

Locator class	describe
NullLocator	No scale
FixedLocator	The scale position is fixed
IndexLocator	Use index as locator （ Such as x=range(len(y)）
LinearLocator	from min To max Command the scale evenly
LogLocator	from min To max Scale by logarithmic distribution
MultipleLocator	Scale and range are cardinal numbers （base） Multiple
MaxNLocator	Find the best position for the maximum scale
AutoLocator	（ Default ） With MaxNLocator Simple configuration
AutoMinorLocator	Locator for minor scale

Format generator class	describe
NullFormatter	There is no label on the scale
IndexFormatter	Set a set of labels as a string
FixedFormatter	Manually label the scale
FuncFormatter	Set labels with custom functions
FormatStrFormatter	Set the string format for each scale value
ScalarFormatter	（ Default ） Set the label for the label value
LogFormatter	Default format generator for logarithmic axes

Explanation of the meaning of the figure and line （ legend ）

function	Parameters	meaning
ax.legend()	There can be no parameters , You can also have the following parameters	Create line meaning
	loc='upper left'	The drawing line shows the position
	frameon=False	Cancel the legend outline
	ncol=2	Number of legend label columns
	fancybox=True	Legend rounded border
	framealpha=0.5	Border transparency
	borderpad=1	Text spacing
	shadow=True	Add shadow

plt.legend()： Create a legend containing each graphic element

x = np.linspace(0,10,1000)
fig, ax = plt.subplots()
ax.plot(x, np.sin(x),'-b', label='Sin')
ax.plot(x, np.cos(x), '--r', label='Cos')

leg = ax.legend(loc='upper left', frameon=True, ncol=2, fancybox=True, framealpha=0.5, borderpad=1, shadow=True)

Select the element shown in the legend

By means of plt.plot() Use or not use label Parameter to determine whether the icon is displayed

x = [1,2,3,4,5,6]
plt.plot(x, [2,5,6,4,2,3], label='1')
plt.plot(x, [3,4,1,6,2,5], label='2')
plt.plot(x, [5,8,4,6,2,9])
plt.plot(x, [2,4,5,8,1,6], label='4')
plt.plot(x, [9,6,4,2,8,3])

#  Show icons 
plt.legend()

Show points of different sizes in the legend

la = np.random.uniform(0,10,100) #  Abscissa 
lo = np.random.uniform(0,10,100)  #  Ordinate 
po = np.random.randint(0,100,100)  #  Color 
ar = np.random.randint(0,1000,100)  #  size 

#  drawing 
plt.scatter(lo, la, label=None, c=po, cmap='viridis', s=ar,linewidth=0, alpha=0.5)
#  Draw a legend 
for ar in [100,200,300]:
    plt.scatter([],[],c='k',alpha=0.3, s=ar,label=str(ar))


#  Show icons 
plt.legend(scatterpoints=1, frameon=False, labelspacing=1)

Configure color bar

Add a title to the color bar

cd = plt.colorbar()
cb.set_label('label')

adopt plt.colorbat Function to create a color bar

#  drawing 
x = np.linspace(0,10,1000)
I = np.sin(x)*np.cos(x[:,np.newaxis])

plt.imshow(I)
plt.colorbar()

Configure color bar

cmap Parameters ： Set the color scheme of the color bar

plt.imshow(I, cmap='gray')

Sequential color scheme ： A color scheme consisting of a continuous set of colors （ for example binary or viridis）
Reciprocal color scheme ： It consists of two complementary colors , Indicates two meanings （ for example RdBu or PuOr）
Qualitative color schemes ： A set of colors in random order （ for example rainbow or jet）

plt.imshow(I,cmap='jet')

Limitation of color bar scale and setting of extended function

It can shorten the upper and lower limits of color values , For data beyond the upper and lower limits , adopt extend Parameters use triangle arrows to represent numbers larger or smaller than the upper limit

x = np.linspace(0,10,1000)
I = np.sin(x)*np.cos(x[:,np.newaxis])
#  Set... For the image 1% noise 
speckles = (np.random.random(I.shape)<0.01)
I[speckles] = np.random.normal(0,3,np.count_nonzero(speckles))

plt.figure(figsize=(10,3.5))

plt.subplot(1,2,1)
plt.imshow(I, cmap='RdBu')
plt.colorbar()

plt.subplot(1,2,2)
plt.imshow(I, cmap='RdBu')
plt.colorbar(extend='both')
plt.clim(-1,1)

Discrete color bar
Sometimes it is necessary to represent discrete data , have access to plt.cm.get_cmap() Parameters

x = np.linspace(0,10,1000)
I = np.sin(x)*np.cos(x[:,np.newaxis])

plt.imshow(I, cmap=plt.cm.get_cmap('Blues', 6))
plt.colorbar()
plt.clim(-1,1)

Manually configure the drawing

#  With a gray background 
ax = plt.axes(fc='#E6E6E6')
ax.set_axisbelow(True)

#  Draw a white grid line 
plt.grid(color='w', linestyle='solid')

#  Hide the lines of the axis 
for spine in ax.spines.values():
    spine.set_visible(False)

#  Hide the upper and right scales 
ax.xaxis.tick_bottom()
ax.yaxis.tick_left()

#  Weaken scale and label 
ax.tick_params(colors='gray', direction='out')
for tick in ax.get_xticklabels():
    tick.set_color('gray')
for tick in ax.get_yticklabels():
    tick.set_color('gray')

#  Set the frequency histogram contour setting and fill color 
ax.hist(x, edgecolor='#E6E6E6', color='#EE6666')

This method is very troublesome to configure , The following method only needs to be configured once and can be used on all graphics

Modify default configuration ：rcParams

import matplotlib as mpl
import matplotlib.pyplot as plt
import numpy as np
from matplotlib import cycler

#fig, ax = plt.subplots()

#  Copy the current rcParams Dictionaries , Change enough to restore 
Ipython_default = plt.rcParams.copy()

#  use plt.rc Function to modify configuration parameters 
colors = cycler('color', ['#EE6666', '#3388BB', '#9988DD', '#EECC55', '#88BB44', '#FFBBBB'])
plt.rc('axes', facecolor='#E6E6E6', edgecolor='none', axisbelow=True, grid=True, prop_cycle=colors)
plt.rc('grid', color='w', linestyle='solid')
plt.rc('xtick', direction='out', color='gray')
plt.rc('ytick', direction='out', color='gray')
plt.rc('patch', edgecolor='#E6E6E6')
plt.rc('lines', linewidth=2)

x = np.random.randn(1000)
plt.hist(x)

#  display picture 
plt.show()

Draw some line drawings to see rc Parameter effect

import matplotlib as mpl
import matplotlib.pyplot as plt
import numpy as np
from matplotlib import cycler

#fig, ax = plt.subplots()

#  Copy the current rcParams Dictionaries , Change enough to restore 
Ipython_default = plt.rcParams.copy()

#  use plt.rc Function to modify configuration parameters 
colors = cycler('color', ['#EE6666', '#3388BB', '#9988DD', '#EECC55', '#88BB44', '#FFBBBB'])
plt.rc('axes', facecolor='#E6E6E6', edgecolor='none', axisbelow=True, grid=True, prop_cycle=colors)
plt.rc('grid', color='w', linestyle='solid')
plt.rc('xtick', direction='out', color='gray')
plt.rc('ytick', direction='out', color='gray')
plt.rc('patch', edgecolor='#E6E6E6')
plt.rc('lines', linewidth=2)

for i in range(4):
    plt.plot(np.random.rand(10))

#  display picture 
plt.show()

stay Matplotlib file There is more information in it

Visual exception handling

The accepted range of certain data is 70±5, I measured that it was 75±10, Is my data consistent with accepted values
In the result of graphic visualization, the error is displayed by graphics , Can provide sufficient information

Basic error line （errorbar）

fmt Parameters ： Control the appearance of lines and points

x = np.linspace(0,10,50)
dy = 0.8
y = np.sin(x)+dy*np.random.randn(50)

plt.errorbar(x,y,yerr=dy,fmt='.k')

errorbar You can define the style of error line graphics

x = np.linspace(0,10,50)
dy = 0.8
y = np.sin(x)+dy*np.random.randn(50)

plt.errorbar(x,y,yerr=dy,fmt='o',color='black',ecolor='lightgray',elinewidth=3,capsize=0)

You can also set the horizontal error （xerr）、 Unilateral error （one-sidederrorbar）、 And other forms of error